Loading market event…
Loading market event…
MajorMarket eventTue, Oct 202 min read
Chinese demand offers an AI hardware sales opportunity, but export permission and delivery determine whether orders become revenue.
ByteDance, Alibaba and Tencent have reportedly placed more than $35 billion in chip orders with Nvidia. These are reported orders, not evidence that the chips have been delivered or that Nvidia has recorded the revenue.[1]
A separate report says the Trump administration halted chip sales to China in 2025 over national security concerns, and that Nvidia subsequently reportedly agreed to give the US government 15% of Chinese chip sales. The supplied reports do not establish a delivery schedule or confirm export permission for these orders.[2][1]
The orders signal that large Chinese technology companies are willing to commit substantial spending to computing hardware. If exports are permitted and deliveries occur, that demand can become Nvidia sales and potentially support the manufacturing and packaging work needed to supply the chips. The order value is not a near-term revenue forecast: timing still depends on permission and fulfillment.[1][2]
Policy also affects the economics of each sale. If the reported 15% government-payment arrangement applies to these orders, Nvidia would not retain all the sales proceeds. For the buyers, receiving the chips could expand computing capacity to train and run AI models, but hardware spending alone does not guarantee sufficient AI service revenue.[2][1]
Fulfilled orders would create sales for Nvidia and could bring more work to its chip manufacturing and packaging suppliers. This is not evidence of new orders across the entire industry; export restrictions and the reported government-payment arrangement still shape the opportunity and its returns.[1][2]
Receiving the chips would let ByteDance, Alibaba and Tencent expand AI computing capacity and support more services for customers. They must fund the hardware first; whether other providers face stronger competition depends on delivery and deployment.[1]